Re: [PATCH]: FTYPINFO command

[Milan Zamazal]

>>>>> "YS" == Yngve Svendsen <address@hidden> writes:

    >> as I already suggested.  As for the return values, I think we
    >> will cope with basically two situations: A single one-line return
    >> value or a list of one-line return values.  So the response
    >> should be in correspondence with the description of the 3xx
    >> responses in the gnatsd command protocol, distinguishing the two
    >> cases by return codes.

    YS> I can't see any reason why we should distinguish between the
    YS> case when there is a single separator and cases when there are
    YS> multiple? The attached set of patches makes gnatsd return a
    YS> string og all the separators, enclosed in 'single quotes'.

I didn't mean just the `separator' property, which I understand as a
single value property, I talked about properties generally, some of them
can truly contain a list of values (one per line in the output).

    YS> The way I implemented this is not as elegant as it should be --
    YS> ideally, I would go for including supported properties for each
    YS> field type in the FieldTypeInfo struct as that would make things
    YS> far cleaner and more extensible. However, I opted for the
    YS> simplest way this time round.

Yes, this was my intention too.  The implementation needn't be elegant
now, the crucial thing is the protocol.

Regards,

Milan Zamazal

-- 
The world is not something you can wrap your head around without needing years
of experience.                              -- Kent M. Pitman in comp.lang.lisp